Analysis Of Human Colonic Pressure Signal Based On Emd

Colonic disease is a disease with high incident, nearly 20% of individuals are affected. However, in the human body organs the understanding of motor function of the colon is still relatively lacking. Colon cavity manometry can accurately reflect the changes of amplitude and frequency, and has been widely used to evaluate the digestive tract motility. But the movement of the human colon is very complicated. The signals are essentially non-stationary, besides since the pressure cavity is often a combination of different dynamic models with different characters and physiological significances, it may have an alias band on the time-frequency domain, which brings difficulty to time-frequency analysis. So traditional time-domain observation or frequency-domain analysis is difficult to give an accurate analysis for the results of manometry.Empirical mode decomposition (EMD) is a new method proposed by Norden E. Huang. It is a local wave decompose method. The complex original signal is decomposed into a finite number of simple components called intrinsic mode functions (IMF). There are two main problems in empirical mode decomposition, endpoint effect and stop conditions. This article uses appropriate method to solve the problems. Through the EMD and power spectral density (PSD) we conclude the criterions of normal and abnormal. We also process colon pressure signal using empirical mode decomposition and information entropy under different conditions. The results proved that we can obtain the characteristic information contained in the colonic pressure signals by this method .